Personal audio devices, including wireless telephones, such as mobile/cellular telephones, cordless telephones, mp3 players, and other consumer audio devices, are in widespread use. Such personal audio devices may include circuitry for driving a pair of headphones or one or more speakers. Such circuitry often includes a speaker driver including a power amplifier for driving an audio output signal to headphones or speakers.
Within both consumer and professional audio equipment, it is often necessary that an operational response among different audio channels is matched. For example, a requirement of such audio systems is that data path group delays and acoustic responses to events (e.g., an interrupt, device error, etc.) be matched across all audio channels, and that at a minimum, such responses be controlled and deterministic.
However, oftentimes in such devices, multiple amplifiers and multiple data paths may be used to drive a plurality of speakers. Most commonly, placement of such amplifiers may be localized to the speakers themselves, meaning that amplifiers may not be in proximity to one another and an amplifier may be subject to local error events not observed by one or more other amplifiers. As an example, a battery connection to a main printed circuit board housing the amplifiers may not be symmetric to the amplifiers, meaning that there may be more series parasitic losses between one amplifier and the battery supply than with another amplifier and the battery supply. If an equivalent amount of power from the battery is consumed by each amplifier, the local supply voltage at one amplifier may be lower than at another amplifier. Such inequality in power consumption may produce a condition whereby one amplifier has to respond to a low voltage error condition, but another amplifier may not observe the error condition.
Often in portable battery powered devices, an amplifier's response of such a low voltage condition may be to reduce the volume of the amplifier, thereby reducing the current consumption of the amplifier in order to preserve the battery's supply voltage for the rest of the system. However, if this audio attenuation is performed on one channel but not another channel, it may have undesirable acoustical effects which create a negative experience to the end user. To illustrate, human hearing is very often sensitive to changes in audio. This sensitivity allows humans to discern things, like direction of the source, approximate distance to the source, whether sound has bounced off an object, and identifying various abnormalities and changes in an expected sound. Both ears are typically leveraged as a part of this process, bringing into the stereo nature of sound. The process of hearing effectively thus creates an “image” of the sound.
For media playback, an audio signal is often intentionally manipulated in order to take advantage of how sound is interpreted by the brain. This intentional audio manipulation is especially true for multi-channel sound. For multi-channel sound, in order for this intentional audio manipulation to be effective and reproducible, certain system-level parameters need to be known and remain constant or deterministic. As electronic audio components become more miniaturized, they also tend to become more distributed in the end system. However, with this distributed architecture, the deterministic requirements, such as channel-to-channel phase and amplitude, have not changed, and thus, an additional approach is needed to ensure the key deterministic relationships of an audio system, such as the multi-device synchronization for audio devices.
Utilizing a software approach, such as an interrupt-based scheme along with control port writes, to mitigate response differences between amplifiers has limitations. Such approach requires a control port master device to: (1) monitor for interrupts, (2) be able to distinguish between different error conditions, possibly requiring control port reads, (3) determine a response to the error condition, and (4) then exchange the information with the other amplifier(s). With conditions continually changing, this approach can be burdensome on the developer, software resources, and still may not have an appropriate response time to properly handle the error condition.
Other unmatched behavior, such as indeterminate group delays (or inter-channel phase shifts), can also produce undesirable acoustic effects. In a controlled manner, manipulating signal phase along with amplitude and frequency shifting can be used to create acoustic effects, such as sound localization, mimicking reverberation, virtual surround sound, etc. However, when a system does not have sufficient control over the systematic behavior of a distributed network of devices, the hardware system can both interact with the multi-channel acoustic performance of the system and create undesirable acoustic effects of its own.
The controlled matching of group delays or system level responses is not just important for purposes of controlling audio output, but is also important for monitoring systems where the signal timing between devices is a key part of the desired functionality. For example, in ultrasonic applications, such as proximity detection or gesture identification, where timing of the signal is critical, a distributed network of analog-to-digital converters connected to ultrasonic microphones would need to be synchronized. Without this synchronization, the detection algorithms might not be able to acquire enough valid data to respond properly.